Let's Pay for Open Source with a Closed-Source Software Levy

This column has often explored ways in which some of the key ideas underlying free software and open source are being applied in other fields. But that equivalence can flow in both directions: developments in fields outside the digital world may well have useful lessons for computing.

Share

This column has often explored ways in which some of the key ideas underlying free software and open source are being applied in other fields. But that equivalence can flow in both directions: developments in fields outside the digital world may well have useful lessons for computing. A case in point is a fascinating post by James Love, Director of Knowledge Ecology International (KEI), a non-governmental organisation concerned with public health and other important issues.

It is called "The value of an open source dividend", and is a discussion of the problems the world of pharma faces because of the distorting effect of patents - problems it shares with the world of computing:

the proliferation of patents, trade secrets and the restrictive licensing of patents, materials, data, know-how and knowledge in general also has negative effects. For example, [the drug company] Gilead recently told KEI that its development efforts for new treatments for hepatitis C were delayed for years by the patent thickets on the hepatitis C virus, a concern predicted by Baruch Blumberg in 1996 in litigation over the overly broad scope of patents on the virus.

In many ways, the drug world provides us with a warning of what could happen in the world of computing if software patents are allowed to become as big a brake on innovation there as they are in the field of medicine. Love goes on:

Sir John Sulston, a joint winner of the 2002 Nobel Prize in Physiology or Medicine, once told KEI that research that is open to everyone is at least nine times more valuable to society than research that is closed. At present, there are very significant financial incentives to make research closed and proprietary, and the asymmetry of incentives leads to an under supply of open research.

Of course, that's also true in the world of software, where there are big financial incentives to keep code closed, even though opening it up would benefit society vastly more than locking it away. But Love describes a very interesting solution to this problem that has been used in the world of drug development:

In 2007, during work by MSF [Médecins Sans Frontières] on a new innovation inducement prize for low-cost point-of-care diagnostics for tuberculosis, concerns about the negative impact of inducement prizes on secrecy were addressed by a proposal for an open source dividend (OSD). The initial proposal was for a percentage of inducement prize money to be allocated to persons who openly shared knowledge, data, materials and technology that was considered significant and useful in the development of the winning diagnostic test.

The idea of offering prizes - typically many millions of pounds - to a company that comes up with an effective treatment for a disease, and then allowing anyone to manufacture that drug, means competition can drive down prices close to costs. That's unlike today, where companies granted monopolies for patented drugs can charge many thousands of pounds per year for them, supposedly to pay for research into the next generation of products (in fact, more is spent by the big pharma companies on marketing than on research.)

Here's how the OSD proposal works in the case of TB diagnostics [.pdf]:

In order to ensure there are incentives for openness and sharing among researchers, the TB Diagnostic Grand Prize money would be divided as follows: the winning entrant would get 90 percent of the prize money; the remaining 10 percent of the prize money would be given to unaffiliated and uncompensated (by the winning entrant) scientists and engineers that openly published and shared research, data materials and technology, on the basis of who provided the most useful external contributions to achieving the end result. This would include research, data, materials and technology that were either placed in the public domain, or subject to open, non-remunerated licenses.

That is, those who put their work into the public domain or under non-restrictive licences could win serious money - even if they don't win the main prize - if their work proved useful in coming up with the winning solution. Love then goes to propose introducing a levy on all drug sales in order to fund this kind of open source dividend on a routine basis, rather than as a one-off, as with drug prizes:

if even 1% of patented drug sales in the United States were set aside into a fund for the OSD, there would be about $2.5 billion each year in rewards to persons who openly shared knowledge, data, materials and technology to drug developers. At 2% of sales, it would be almost $5 billion each year. The availability of the OSD would revolutionize university licensing policies, and also the policies of businesses that were holding as secrets research that was no longer under development.

Most interestingly for readers of this column, he goes on:

The new OSD market incentives to share knowledge could easily be extended to software development, with even small fees on the sale of licensing of proprietary software, or devices that used software, such as computers or mobile computer devices.

That's an incredibly bold, and very attractive idea: using a small levy on all closed-source software sales to fund the development of open source code, which would then be freely shared. Of course, some thought would need to be given to how that would work in practice, and how cloud-based services might be included, but those are just details. Love points out why such a levy would make economic sense:

The companies that sold ... mobile computing devices would arguably benefit the most, if the OSD induced more royalty-free sharing of inventions, data, materials and source code for software. This is because it would lower the costs of licensing these inputs, and also because the overall products would improve at a faster rate, creating more valuable products for end users.

That's a hugely important point. When code is made available under an open source licence, it allows *all* companies to use it without payment. That not only saves them money, but enables more innovative products to be delivered to the market, and faster, so the public wins too.

Of course, the idea of taxing closed-source software in order to promote open innovation in this way would be met by howls from certain companies - and a vast amount of lobbying against the plan. Nonetheless, the rapid rise of the thriving global ecosystem around Android shows just how effective such an open approach can be in attracting investment and creating popular new products. Maybe it's an idea whose time has come.  Let's start spreading the word.

Follow me @glynmoody on Twitter or identi.ca, and +glynmoody on Google+

"Recommended For You"

The Threat Of Parallel Filing A New Chapter for Open Source?